— 215 — 



Prophasis. 



In synapsis and the contracted spirem-st age many 

 parallel windings or loops can be seen (fig. 1 — 3), but no arrange- 

 ment in pairs can be detected. In most places the threads of the 

 coil present a structure like a single string of beads. 



The diakinesis begins with the forming of rings, and 

 oo - shaped figures (fig. 4 — 7). The double-chromosomes in the be- 

 ginning resemble dumb-bells bent together in U-form, and the 

 chromatin swellings in each end of them represent the single 

 components of the double-chromosome (fig. 8 — 10). Later on the 

 components approach and in the typical diakinesis they lie just 

 beside each other. 



In Miyaji's paper (1913) some drawings, apparently re- 

 presenting the diakinesis in different Viøfø-species, exhibit gemini 

 resembling those I have seen in V. arvensis, but the drawings 

 only show the outline of the gemini. 



Metaphasis and Anaphasis. 



The heterotypical metaphasis in Viola arvensis is some- 

 what irregular. (Instances: figs. 11 and 12). The chromosomes 

 seem to have difficulties in separating. Exact counting is impos- 

 sible here. Figs. 11a and b show the chromosomes to be more 

 than 30 at any rate. An irregular Anaphasis has been drawn 

 in fig. 13 a and b. Four smaller chromosomes not following the 

 others are lying in the strands of linin between the pole plates, 

 apparently hesitating between the two courses open to them. 

 (The three isolated chromosomes to the left in fig. 13 a probably 

 have been displaced in cutting). The homoeotypical division 

 usually goes on very regularly as shown in fig. 14 and figs. 17 — 22. 



Vegetative Mitoses. 



In the tapetum cells of the anthers fine vegetative mitoses 

 can often be seen. It is rather difficult to count the chromosomes 

 here. They are long, often lying in prolongation of each other 

 and are of course rather numerous. Fig. 25 shows such a mitosis 

 from a Viola arvensis in the equatorial plate stage. About 60 

 chromosomes can be seen. The division of the chromosomes must 

 have taken place, and the somatic number of chromosomes must 

 be about 30. This is sufficient to show that a reduction takes 



